Donor antigen-specific regulatory T-cell function affects outcome in kidney transplant recipients

Transcription

1 & 211 International Society of Nephrology original article Donor antigen-specific regulatory T-cell function affects outcome in kidney transplant recipients Igor Tsaur 1, Martin Gasser 2, Beatriz Aviles 2, Jens Lutz 3, Lydia Lutz 3, Martin Grimm 2, Volkmar Lange 2, Kai Lopau 4, Uwe Heemann 3, Christoph-Thomas Germer 2, Anil Chandraker 5,6 and Ana Maria Waaga-Gasser 2,6 1 Department of Urology, University of Frankfurt, Frankfurt, Germany; 2 Department of Surgery I, University of Wuerzburg, Wuerzburg, Germany; 3 Department of Nephrology, Klinikum rechts der Isar der Technischen Universität München, Munich, Germany; 4 Department of Internal Medicine, Nephrology, University of Wuerzburg, Wuerzburg, Germany and 5 Transplantation Research Center, Renal Division, Brigham and Women s Hospital, Harvard Medical School, Boston, Massachusetts, USA Chronic transplant dysfunction, a major impediment to long-term allograft survival, is caused by several factors including an ongoing alloimmune response termed chronic rejection. To define some of these factors further, we selected 17 patients mismatched to their donors from 623 patients transplanted at a single center. Patients were categorized according to their immunosuppressive treatment and further divided into those with stable or chronic allograft dysfunction. Donor human lymphocyte antigen allopeptide-specific T-cell lines were then generated from stable patients and those with biopsy-proven chronic allograft nephropathy. Increased amounts of CD4 þ CD25 þ regulatory T cells (Tregs) and Tregassociated gene expression profiles were found in cell lines derived from the patients with stable compared with those with chronic allograft dysfunction. Furthermore, a higher percentage of Tregs was found in patients with stable graft function on tacrolimus-based compared with cyclosporinebased immunosuppression protocols. Patients with stable graft function had a significantly higher expression of interleukin (IL)-4 and IL-1, whereas the cytokines IL-2, IL-17, and interferon-c were significantly higher in patients with allograft dysfunction in vitro. Thus, enhancing the operational role of naturally occurring donor-specific Tregs in allograft recipients by adjusting the immunosuppression protocol may be advantageous particularly for patients with ongoing chronic rejection. Kidney International (211) 79, ; doi:1.138/ki ; published online 26 January 211 KEYWORDS: chronic allograft dysfunction; kidney transplantation; mycophenolate mofetil; regulatory T cells; tacrolimus Correspondence: Ana Maria Waaga-Gasser, Department of Surgery I, University of Wuerzburg, Oberduerrbacher Street 6, Wuerzburg 978, Germany. waaga-gasser@chirurgie.uni-wuerzburg.de 6 These authors contributed equally to this work. Received 16 February 21; revised 4 October 21; accepted 28 October 21; published online 26 January 211 During the past two decades, the introduction of the calcineurin inhibitors cyclosporine A (CsA) and tacrolimus has significantly improved 1-year kidney allograft survival. However, long-term graft function often remains suboptimal because of the development of chronic allograft dysfunction (CAD). Chronic rejection has an important role in the development and progression of chronic graft damage. The alloimmune response is influenced by a specialized subset of regulatory T cells (Tregs) that participate in the maintenance of immune homeostasis. 1,2 Tregs can modify rejection episodes as they are able to downregulate activated allospecific T cells. 1 3 It has also been shown that Tregs can actively maintain hyporesponsiveness to alloantigens in renal transplant patients 4,5 and that calcineurin-dependant signaling has an important role in the expansion of Tregs. 6 So far, only few data have been reported with respect to the effect of the different calcineurin inhibitors tacrolimus and CsA on the prevalence of donor antigen-specific Tregs after kidney transplantation. Furthermore, the association between graft function and the prevalence of Tregs in the graft recipient is not clear. A quantitative effect of calcineurin inhibitors on Tregs could not be detected in kidney and lung transplanted patients; 7,8 however, an effect of these drugs could be detected in CD8 þ CD28 T cells. 8 There is some evidence that tacrolimus-based immunosuppressive protocols might preserve graft function better than CsA-based protocols, 9 possibly because tacrolimus might better control subacute immune responses in the allograft, thus reducing chronic rejection more effectively than CsA. The relative effect of the different calcineurin inhibitors on Tregs may explain the different efficacies of tacrolimus compared with CsA. This is supported by observations demonstrating that tacrolimus had a protective effect on Tregs in atopic dermatitis. 1 So far, only limited data are available characterizing Tregs after organ transplantation, particularly with respect to the immunosuppressive protocol. Kidney International (211) 79,

2 original article I Tsaur et al.: Donor antigen-specific regulatory T-cell function RESULTS Patients From a total of 17 patients selected for this study, a greater number of patients on tacrolimus (Tac; Tac/mycophenolate mofetil (MMF)/steroids) had stable graft function (n ¼ 2/ 3, 66.7%) as compared with patients receiving CsA either Stable CsA/ster 63 Patients Patients Cell lines CsA/MMF/ster DR mismatched Tac/MMF/ster CAD Stable CAD Stable CAD CsA, Cyclosporin A; MMF, mycophenolate mofetil; stable, stable renal function; ster, steroids; Tac, tacrolimus; creatinine 1.6 mg/dl; CAD, chronic allograft dysfunction; creatinine >1.6 mg/dl; Performed cell lines Figure 1 Study patients flow chart. Of 63 transplanted patients, 17 were human lymphocyte antigen DR mismatched with their donors for the available donor-specific peptides and selected for further analysis. The patients were categorized according to their creatinine levels. Patients with a creatinine p1.6 mg/dl were considered to have stable graft function (stable) and those with a creatinine 41.6 mg/dl and biopsy-proven chronic allograft nephropathy as chronic allograft dysfunction (CAD). T-cell lines were generated from patients with stable function or biopsy-proven CAD. with or without MMF (CsA/MMF/steroids, n ¼ 21/47 (44.7%); CsA/steroids, n ¼ 11/3 (36.7%)). Of the 55 patients classified as having CAD, a total of 24 (43.6%) patients were biopsied and had evidence of transplant glomerulopathy, vasculopathy, interstitial fibrosis, and when detectable, tubular atrophy. A total of 19 cell lines were generated from these patients with biopsy-proven chronic rejection (Figure 1). Eighteen cell lines were generated from matched stable transplant patients. Three out of five patients in the CsA/steroid group, no patients in the CsA/MMF/steroid group, and 2/16 patients in the Tac/MMF/steroid group received anti-t-lymphocyte globulin-induction therapy, but there were no differences with respect to age, gender, human lymphocyte antigen (HLA) mismatch, serum creatinine, number of acute rejections, and cold ischemia time and the development of chronic allograft dysfunction. Twelve hour trough tacrolimus and CsA serum levels are presented as mean values (ng/ml) and range in both representative groups (Tac/MMF/steroid and CsA/MMF/steroid). For stable patients, the mean Tac level at 24 months was 11±2.7 ng/ml (range ) and for CAD patients at 1.3±2.6 ng/ml (range ). The mean CsA level at 24 months for stable patients was 142.7±34.5 ng/ml (range ) and for CAD patients 135.2±37.7 ng/ml (range ; Table 1). Comparison of graft function in the different groups of patients Serum creatinine levels were similar throughout the observation period in stable patients on tacrolimus- and CsA-based immunosuppression. Patients with CAD receiving CsA-based immunosuppressive therapy had increasing creatinine levels, Table 1 Demographic data of the patients from whom T-cell lines were generated (24 months after Tx) CsA/steroids (n=5) CsA/MMF/steroids (n=16) Tac/MMF/steroids (n=16) Stable CAD Stable CAD Stable CAD Recipient demographics Number of recipients (n=17) Age (mean±s.d. in years) ± ± ± ±7.1 Gender (%) Female Male HLA mismatch (number of mismatches of six HLA Ag, in mean±s.d.) ±.9 3.1± ±.9 3.1±1.5 Serum creatinine (mean±s.d. in mg/dl) ±.2 a 2.5±.2 a.9±.1 b 2.2±.3 b CsA/Tac trough level (mean±s.d. in mg/dl) and range ND ND 142.7± ± ± ± Proteinuria 41 g/24 h (number of patients/group) Acute rejection (number of patients/group) Biopsy (Banff classification of chronic allograft nephropathy or description) Donor characteristics Type of donor (number/group) Deceased Living Cold ischemia time (mean±s.d. in hours) ± ± ± ±5.8 Abbreviations: CAD, chronic allograft dysfunction; CsA, cyclosporine A; HLA, human lymphocyte antigen; MMF, mycophenolate mofetil; ND, not determined; stable, stable renal function; Tac, tacrolimus; Tx, transplantation. a CsA/MMF/ster versus Tac/MMF/ster Po.5. b CsA/MMF/ster versus Tac/MMF/ster CAD, Po Kidney International (211) 79,

3 I Tsaur et al.: Donor antigen-specific regulatory T-cell function original article whereas patients on tacrolimus had quite constant creatinine levels during the observation period (Figure 2); however, none of the grafts were lost by 6 months. Cytokine analysis in the serum of transplanted patients In both the sera of patients from the CsA/MMF/steroid and Tac/MMF/steroid groups, cytokine expression profiles were measured at 24 months, the time point when T-cell lines were generated, and at 6 months after transplantation. Generally, patients with CAD expressed more interleukin (IL)-2, IL-2R, IL-17, and interferon (IFN)-g than patients with stable graft function (Figure 3). Interestingly, the patients with stable function expressed more IL-4 and IL-1 in both groups. We observed that patients on Tac/MMF/steroids with stable graft function expressed a higher amount of IL-1 than those on CsA/MMF/steroids at 24 and 6 months (Po.5). In addition, the serum concentrations of IL-2 and IFN-g were higher in patients on CsA/MMF/steroids with CAD than those on Tac/MMF/steroids at 24 and 6 months time points (Po.1). Flow cytometric analysis of CD4 þ CD25 þ T cells Higher percentages of CD4 þ CD25 þ T cells were observed in T-cell lines generated from patients with stable graft function as compared with those generated from patients with CAD in both groups (CsA/MMF/steroids and Tac/ Creatinine (mg/dl) 4 CsA+steroid CsA+MMF+steroid Tac+MMF+steroid CAD 3 Stable Months Creatinine (mg/dl) Months Months Figure 2 Serum creatinine after transplantation. Creatinine levels were similar during the observation period in stable patients (stable) on tacrolimus- (Tac/MMF/ster) and CsA (CsA/MMF/ster)-based immunosuppression. Patients with chronic allograft dysfunction (CAD) receiving CsA tended to have increasing creatinine levels over time as compared with patients on tacrolimus. Creatinine was measured on day and 12, 24 (generation of T-cell lines, CsA/MMF/ster versus Tac/MMF/ster: 2.6±.5 versus 2.3±, Po.1), 48, and 6 months after transplantation. Each dot represents one patient; the horizontal lines mark the mean value of the measured creatinine concentration. CsA, cyclosporine A; MMF, mycophenolate mofetil; ster, steroid; Tac, tacrolimus CsA/MMF/ster Concentration in serum (pg/ml) Tac/MMF/ster P <.1 CAD 24 mo (n =8, generation of T-cell lines) P <.1 Stable 24 mo (n =8, generation of T-cell lines) CAD 6 mo (n =8) Stable 6 mo (n =8) P <.1 P <.1 P <.5 P <.1 IL-2 IL-2R IL-4 IL-1 IL-17 IFN-γ P <.1 Concentration in serum (pg/ml) P <.1 P <.1 P <.5 P <.1 P <.5 IL-2 IL-2R IL-4 IL-1 IL-17 IFN-γ Figure 3 Cytokine expression in the serum of patients at 24 and 6 months after transplantation. Patients with chronic allograft dysfunction (CAD) expressed significantly more interleukin (IL)-2, IL-2R, IL-17, and interferon (IFN)-g than patients with stable renal function. Patients with stable graft function expressed more IL-4 and IL-1, irrespective of the immunosuppressive protocol. The P-values indicate the significant differences between patients with CAD and stable renal function 24 months after the transplantation (the time point when the T-cell lines were generated) and 6 months after the transplantation. CsA, cyclosporine A; MMF, mycophenolate mofetil; ster, steroid; Tac, tacrolimus. Kidney International (211) 79,

4 original article I Tsaur et al.: Donor antigen-specific regulatory T-cell function Table 2 Characterization of T-cell lines generated from patients with chronic allograft dysfunction and stable renal function Cell lines Phenotype (%) HLA mismatch CD3+CD8+ CD3+CD4+ CD4+CD25+ CsA/MMF/ster CAD 1 DR DR DR DR DR DR DR DR Mean±s.d. 8.± ±4.2 1.±.6 Stable 1 DR DR DR DR DR DR DR DR Mean±s.d. 5.9± ± ±.9 Tac/MMF/ster CAD 1 DR DR DR DR DR DR DR DR Mean±s.d. 7.4± ±4.2 8.±1.4 Stable 1 DR DR DR DR DR DR DR DR Mean±s.d. 4.6± ± ±1.1 Abbreviations: CAD, chronic allograft dysfunction; CsA, cyclosporine A; HLA, human lymphocyte antigen; MMF, mycophenolate mofetil; stable, stable renal function; ster, steroid; Tac, tacrolimus. MMF/steroids, Po.1). In addition, the proportion of CD4 þ CD25 þ was significantly increased in T cell lines generated from patients with stable graft function on Tac/MMF/steroid protocol as compared with stable patients on CsA/MMF/steroids (Po.1; Table 2). Real-time PCR analysis of Foxp3 expression in PBMCs from patients with stable graft function and on tacrolimus protocol To ascertain whether the high levels of CD25 expression in CD4 þ T cells in tacrolimus-treated patients reflected a regulatory phenotype, we evaluated mrna expression of Foxp3 (refs. 6, 11, 12). Peripheral blood mononuclear cells (PBMCs) from the patients with stable graft function on Tac/ MMF/steroids had a 2- to 4.5-fold higher gene expression of CD4, CD25, CTLA-4, Foxp3, and IL-1 as compared with patients treated with CsA/MMF/steroids and stable graft function (Figure 4). No significant differences in expression of cytokines were observed between tacrolimus- and CsAtreated patients with chronic allograft dysfunction, except for the expression of IFN-g, which was higher in the CsA-treated patients. The expression of transforming growth factor (TGF)-b was significantly higher in CAD patients, the difference between Tac- and CsA-based regimen was not statistically different. Morphological and functional characterization of T-cell lines All T-cell lines were generated (24 months after transplantation) by repeated stimulation (3 times) of PBMCs from patients with stable graft function and patients with CAD with the appropriate mismatched donor-derived HLA-DR peptides (Table 2). The proliferative response of the T-cell lines to the mismatched donor-derived and irrelevant (control) HLA-DR peptides was tested. T-cell lines generated from patients with CAD had a higher proliferative response as compared with those from patients with stable graft function (Figure 5). No significant proliferation above base line was observed after stimulation with irrelevant HLA-DR control peptides in the T-cell lines generated from either patient group (data not shown). CD4 þ and CD8 þ T cells were present in all the T-cell lines independent of the immunosuppressive protocol. T-cell lines generated from PBMCs of patients with stable graft function contained higher amounts of CD4 þ CD25 þ cells, as compared with patients with CAD (Table 2). Patients with stable function treated with tacrolimus had a higher proportion of CD4 þ CD25 þ cells as compared with patients on CsA with stable function (Table 2), suggesting a higher proportion of Tregs in patients treated with Tac/MMF/steroids. Regulatory function of T cells from patients with stable graft function We next investigated whether the T-cell lines generated from patients with stable graft function were able to regulate the alloimmune responses of T cells from patients with CAD. We set up a co-culture system in which T-cell lines generated from patients with stable graft function were irradiated (to prevent their proliferation) and incubated with T cells from patients with CAD together with the relevant HLA-DR allopeptides in vitro. 13,14 T-cell lines generated from patients with CAD and on tacrolimus protocol showed a higher proliferative response to their specific HLA-DR allopeptide, as compared with the T-cell lines generated from patients with stable graft function. The addition of irradiated T-cell lines from the stable patients significantly inhibited the proliferative response of the T-cell lines derived 18 Kidney International (211) 79,

5 I Tsaur et al.: Donor antigen-specific regulatory T-cell function original article CsA/MMF/ster CAD (n = 8) Tac/MMF/ster CAD (n = 8) P <.1 CsA/MMF/ster stable (n = 8) Tac/MMF/ster stable (n = 8) 1 P <.1 9 P <.1 8 P <.1 P <.1 7 P <.1 P < CD4 CD25 CTLA-4 Foxp3 IL-1 IFN-γ TGF-β Figure 4 Gene analysis of peripheral blood mononuclear cells (PBMCs) from patients at 24 months after transplantation. PBMCs from patients with stable graft function (stable) on tacrolimus had a 2- to 4.5-fold higher gene expression of CD25, CTLA-4, Foxp3, and interleukin (IL)-1 than those on CsA (Tac/MMF/ster versus CsA/MMF/ster, n ¼ 8 patients/group). Fold difference is expressed as 2 DDCt. CsA, cyclosporine A; MMF, mycophenolate mofetil; ster, steroid; Tac, tacrolimus. χ-fold difference c.p.m. ( 1 3 ) CsA/MMF/ster P <.1 P <.1 c.p.m. ( 1 3 ) Tac/MMF/ster P <.1 P <.1 CAD/pep Stable/pep Irr. stable/pep CAD/irr. stable/pep CAD/irr. stable/pep/ anti-il-1-mab stable/pep/ anti-il-1-mab Figure 5 Regulatory function of T-cell lines derived from patients with stable graft function. Representative proliferation assay of a co-culture of a T-cell line generated from patient 3 with stable graft function on CsA and patient 2 on tacrolimus (Table 2) irradiated with 3 rad, cultured in a 1:1 ratio with T-cell line derived from patient 6 with chronic allograft dysfunction (CAD) on CsA and patient 6 on tacrolimus (Table 2). From left to right (1) T-cell line from CAD patient specific for HLA-DR1 (42 62) þ peptide HLA-DR1 (42 62); (2) T-cell line generated from patient with stable graft function specific for HLA-DR1 (42 62) þ peptide HLA-DR1 (42 62); (3) same T-cell line but irradiated; (4) CAD T-cell line specific for HLA-DR1 (42 62) þ irradiated T-cell line derived from patient with stable function specific for HLA-DR1 (42 62) þ peptide HLA-DR1 (42 62); (5) T-cell line from CAD patient specific for HLA-DR1 (42 62) þ irradiated T-cell line derived from patient with stable graft function specific for HLA-DR1 (42 62) þ peptide HLA-DR1 (42 62), and neutralizing IL-1 mab; (6) T-cell line from patient with stable graft function specific for HLA-DR1 (42 62) þ peptide HLA-DR1 (42 62) þ neutralizing interleukin (IL-1) mab. Data are expressed as mean±s.d., repeated three times. Representative example for n ¼ 8 T-cell lines/group. CsA, cyclosporine A; HLA, human lymphocyte antigen; mab, monoclonal antibody; MMF, mycophenolate mofetil; pep, peptide; ster, steroid; Tac, tacrolimus. from CAD patients (53.3%; Figure 5). These effects were less pronounced in T-cell lines from patients on CsA (38.7%; Figure 5). To further dissect the mechanisms of regulation, we measured the proliferation by the T-cell lines in the presence or absence of neutralizing IL-1 monoclonal antibody (mab) or isotype-matched control. Incubation with neutralizing IL-1 mab enhanced the proliferative response of the co-cultured T cells, as compared with the response without IL-1 mab. Interestingly, incubation of T cells generated from stable patients with neutralizing IL-1 mab increased the proliferative response, as compared with the response of those cells without neutralizing IL-1 mab. Our data demonstrate a regulatory function of HLA-DR allopeptidespecific T-cell lines derived from patients with stable graft function in human transplant patients and suggest that this is at least in part dependent on IL-1 production. DISCUSSION From several clinical studies it is known that patients on tacrolimus-based immunosuppression had lower serum creatinine values and higher glomerular filtration rates, as compared with patients on CsA-based immunosuppression. 9,15 The reason for this effect is not entirely clear yet. In our study, we also observed better renal function in a greater proportion of patients on an immunosuppressive protocol consisting of Tac/MMF/steroids, as compared with patients on CsA/MMF/steroids. However, it has been demonstrated that chronic rejection is an important contributor to worsening renal allograft function. 16 It is believed that tacrolimus more effectively controls the alloimmune response as compared with CsA, thus inhibiting subacute as well as chronic rejection, leading to improved graft function and potentially longer graft survival. This is supported by the observation of lower rates of acute rejection Kidney International (211) 79,

6 original article I Tsaur et al.: Donor antigen-specific regulatory T-cell function in patients treated with tacrolimus as compared with CsA. 9 Whether this is related in part to the fact of immune regulation through CD4 þ CD25 þ Tregs is still unknown. A large body of evidence supports the fact that CD4 þ CD25 þ Treg cells have potent immune regulatory properties in vitro 6,13,17 2 and are able to downregulate an alloantigen-directed immune response. 21,22 Recent studies have shown that the frequency of circulating CD4 þ CD25 þ Tregs was associated with a better graft function in patients after lung and kidney transplantation. 23,24 However, little is known about the effect of the two different calcineurin inhibitors on CD4 þ CD25 þ Treg generation. CD25 is expressed on a variety of T cells including effector and memory T cells that respond aggressively to alloantigens. 25 In this study, we demonstrate that isolated T-cell lines from patients with stable graft function have higher mrna levels of CD4, CD25, and Foxp3 as well as higher numbers of T cells of the CD4 þ CD25 þ regulatory phenotype, as compared with cell lines from patients with chronic allograft dysfunction. Thus, the increased levels of Foxp3 together with high levels of CD25 expression in CD4 þ T cells appear to reflect a Treg phenotype in our patients. This is supported by the results of our functional analysis. The proliferation of T-cell lines from patients with CADwas inhibited after co-culture with T-cell lines from patients with stable graft function in the presence of donor-specific HLA-DR peptide. These results are in accordance with our previous study in an experimental animal model. In that study we demonstrated that animals injected with Th2-type cell clones generated from tolerant animals achieved donor-specific tolerance to the renal allograft, as evidenced by the acceptance of donor-specific skin allografts. Enrichment of Tregs producing Th2-type cytokines were also observed in the target organs by immunohistology. Animals, which received Th1 cell clones (derived from animals with acutely rejected allografts), rejected their allografts in an accelerated manner. This suggests that the generation of Tregs can occur after transplantation and their effect may be demonstrated, if the aggressive alloimmune response is controlled, as illustrated in our previous animal study through the use of a low-dose CsA protocol. That means that faced with a reduced aggressive alloreactive T-cell response, Tregs are able to emerge as the dominant T-cell population and control the immune response to the allograft. 26 T-cell interactions through production of cytokines such as IL-1 and competition for antigen-presenting cells are likely to be important components of the molecular interactions critical to both linked suppression and infectious tolerance. The evidence of enrichment of Tregs, producing Th2 type cytokines, in allografts from animals injected with Th2 cell clones, suggests that Tregs operate, at least in part, by locally protecting the tissues against an allospecific attack. The results also add to our previous pilot study in transplant patients treated with CsA. 11 How does tacrolimus- and CsA-based immunosuppressive therapy influence the number and function of Tregs? In other studies, calcineurin inhibitors were associated with lower numbers of Tregs in comparison to a sirolimus-based immunosuppression protocol. 6,27,28 We observed higher numbers of CD4 þ 25 þ Foxp3 þ Tregs in patients treated with tacrolimus as compared with those taking CsA. These results are in congruence with the co-culture experiments that demonstrated a more profound inhibition of the proliferative response of T-cell lines from patients with chronic graft dysfunction by T-cell lines from patients with stable function on tacrolimus, as compared with those on CsA. Further evidence of a functional role of Tregs is derived from the inhibitory effects on Treg function of a neutralizing IL-1 antibody in the co-culture experiments. It can be speculated that this effect might participate in the improved graft function of patients treated with tacrolimus as compared with CsA. Gene expression analysis resulted in significantly different levels of CD25 and Foxp3 in patients on a tacrolimus protocol as compared with CsA-based immunosuppression, only if stable graft function was present. With regard to the frequency of Tregs, these results are in accordance with studies in humans that demonstrated the lowest rate of acute rejections and better graft function in patients on a Tac/MMF/steroid-based immunosuppressive protocol. 9 Patients with stable graft function had a significantly higher expression of IL-4 and -1, whereas the cytokines IL-2, IL-17, TGF-b, and IFN-g were significantly higher in patients with allograft dysfunction. Interestingly, no increasing in TGF-b in stable patients was observed, indicating that the immunoregulation in our patients was TGF-b independent. There were no significant differences with respect to cytokine expression between patients receiving tacrolimus or CsA. This could point to a higher activity of Tregs in patients with stable graft function. The higher levels of IFN-g in patients with decreased graft function could be indicative of an ongoing chronic rejection process within the allograft. Taken together, T-cell lines from patients on a Tac/MMF/ steroid protocol had a comparably higher percentage of cells with morphological/molecular characteristics of Tregs than those on CsA/MMF/steroid. Thus, Tac/MMF/steroidbased immunosuppression seems to be accompanied with an increased amount of Tregs in comparison to CsAbased immunosuppressive protocols in kidney transplant recipients. PATIENTS AND METHODS Patients From a total of 63 kidney transplant recipients, 17 patients who were DR mismatched for one of the five candidate HLA- DR antigens (DR1, DR2, DR3, DR4, and DR7) were included into the study. All patients were recipients of deceased donor allografts. Patients were treated with one of the three following immunosuppressive protocols: cyclosporine A þ steroids (CsA/ster; n ¼ 3), CsA þ MMF þ ster (CsA/MMF/ ster; n ¼ 47), or tacrolimus þ MMF þ ster (Tac/MMF/ster; 11 Kidney International (211) 79,

7 I Tsaur et al.: Donor antigen-specific regulatory T-cell function original article n ¼ 3). Patients who had a serum creatinine consistently below 1.6 mg/dl by month 24 following transplantation were considered to be stable, whereas patients with a consistently elevated serum creatinine above 1.6 mg/dl at month 24 after transplantation were assigned to the group with CAD group. Within the CAD group, T-cell lines were generated only from patients with biopsy-proven chronic allograft injury (Figure 1). Matched patients by age, gender, and as much as possible by HLA were selected from the stable group, and T- cell lines were generated from these patients (Figure 1). All patients were followed regularly at 3- to 6-month intervals, resulting in a completeness index (that is, observed divided by expected follow-up time) of.96. The follow-up time post-transplant was 6 months. Creatinine levels (mg/dl) were measured in patients from whom T-cell lines were generated (day ¼ day of transplantation, 12 months after transplantation, 24 months ¼ generation of T-cell lines, 48, and 6 months). Blood samples were collected after obtaining informed consent. The study protocol was approved by the local medical ethics committee. HLA-DRB peptides A panel of peptides was synthesized corresponding to the full-length b-chain hypervariable regions of HLA-DRB1 11, 151, 31, 41, and 71 (ProImmune, Littlemore, UK), as previously reported. 11 The peptides used for this study included the HLA-DRB11 (HLA-DR1 residues 6 21, 22 41, 42 62, 63 8, 81 94), HLA-DRB151 (HLA-DR2 residues 1 2, 21 4, 41 6, 61 8, 81 94), HLA-DRB31 (HLA-DR3 residues 6 21, 22 41, 42 62, 63 8, 81 94), HLA-DRB41 (HLA-DR4 residues 1 18, 21 42, 62 8), and HLA-DRB71 (HLA-DR7 residues 1 18, 2 41, 5 65, 62 82). Establishment of HLA-DRB-specific T-cell lines PBMCs from patients with stable graft function and from CAD were collected 24 months after transplantation and were separated by Ficoll-Hypaque (Amersham Biosciences AB, Freiburg, Germany) density gradient centrifugation. The cells were then washed twice in RPMI 164 medium (Gibco, Eggenstein-Lopoldshafen, Germany) containing 4% normal human serum (Gibco), 1 U/ml penicillin, 1 mg/ml streptomycin, 5 mmol/l HEPES, 1% nonessential amino acids, and 1 mmol/l sodium pyruvate (Gibco), and spun at 4 g. The cells were counted after resuspension in RPMI medium that was the same as above, except containing 1% normal human serum. PBMCs from human renal allograft recipients (4 1 6 ) were cultured separately with 5 mg/ml of the mismatched donor-derived HLA-DR allopeptides. All T-cell lines were incubated at 371C with 5% CO 2 for 4 days. The cells (2 1 5 ) were repeatedly stimulated with the donor-specific allopeptides at 7-day intervals in the presence of irradiated (3 Gy) PBMCs from the recipient as antigenpresenting cells. During the remaining period, the T-cell lines were cultured in complete medium with 1% human T-Stim culture supplement without phytohemagglutinin (BD Biosciences Pharmingen, Heidelberg, Germany). T-cell lines were generated after three cycles of stimulation. Luminex analysis The human cytokine bead assay kit (Luminex, Biosource, Nivelles, Belgium) was used for measurement of IL-2, IL-2R, IL-4, IL-1, IL-17, and IFN-g. The assay was performed according to the manufacturer s instruction (Biosource). Briefly, the filter plate was prewetted with 2 ml of working solution. The beads (25 ml) were pipetted into each well. Sera or supernatants of the T-cell lines after stimulation with the peptides (proliferation assay) or standards (5 ml/each) and supernatants with the neutralizing anti-il-1 mab or rat immunoglobulin-g1 isotope control were pipetted into the wells and incubated with the beads for 2 h. After washing, biotin-conjugated detector antibody was added and incubated for 1 h, followed by an incubation of 3 min with R-phycoerythrin-conjugated streptavidin. Analysis was carried out in the Luminex 1 instrument (Gurce, Nivelles, Belgium). The results were expressed as pg/ml. Real-time polymerase chain reaction mrna expression of the representative surface molecules and cytokines (CD4, CD25, CTLA4, Foxp3, IL-1, IFN-g, and TGF-b) (MWG Biotech, Ebersberg, Germany) was analyzed in the T-cell lines generated from the recipients by Real Time PCR assay. The primers used were purchased from Sigma Genosys (Woodlands, TX). The glyceraldehyde-3-phosphate dehydrogenase (Sigma Genosys) mrna levels were analyzed as housekeeping gene expression for relative quantification and complementary DNA quality control. Reverse transcription from RNA to complementary DNA was carried out by using iscript cdna Synthesis Kit (Bio-Rad, Hercules, CA). Initial denaturation at 951C for 15 min was followed by 39 cycles of a denaturation step at 951C for 15 s, an annealing step at 57.51C for 3 s, and an extension step at 721C for 3 s. The quantification data were analyzed with the LightCycler analysis software (My Research, Biozym, Oldendorf, Germany). Reproducibility was confirmed by independent PCR repeated twice. The average threshold cycle (C t ) value was calculated as the cycle number at which the fluorescence of the reporter gene reaches a fixed threshold. The difference (DC t ) between the average C t values of the samples in the target wells and those of the housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase, was assessed, followed by the calculation of the difference between the average DC t values of the PBMCs of the patients or T-cell lines for each target and the DC t value of healthy volunteers for that target (DDC t ). The relative quantification value, fold difference, is expressed as 2 DDC t. Proliferation assay After generation of the T-cell lines, cells were cultured with 1 mg (this dose was used based on our previous publication showing the dose response to this peptide) 11 of the relevant donor-mismatched HLA-DR allopeptide in Kidney International (211) 79,

8 original article I Tsaur et al.: Donor antigen-specific regulatory T-cell function 96-well U-bottom plates (VWR, Darmstadt, Germany). Human neutralizing antibody, rat anti-human IL-1 mab (.5 25 mg/ml), or rat immunoglobulin-g1 isotope control antibody (all from PharMingen, San Diego, CA) were added to the cultures. As the optimal effects of anti-il-1 mab were observed at a concentration of 1 mg/ml, the same dose was used for anti-il-1 and isotope control antibodies in the experiments. Proliferation was measured by 3 [H]-thymidine (1 mci) (Amersham Biosciences Europe, Freiburg, Germany) incorporation. All assays were set up in quadruplicates, and the results were expressed as c.p.m.±s.e.m. In the co-culture system, irradiated (3 Gy) T-cell lines generated from patients with stable allograft function and on tacrolimus or CsA immunosuppression (2 1 5 ) were incubated with an equal number of the T-cell lines (ratio 1:1) generated from patients with CAD and on tacrolimus protocol and the relevant HLA- DR allopeptide. Background proliferation was assessed by incubating the cells with culture media alone without the peptides. Flow cytometry (fluorescence-activated cell sorter) analysis T-cell lines (2 1 5 cells/sample) were stained with conjugated ECD-anti-human CD3, PC5-anti-human CD4, ECD-anti-human CD8, and PE-anti-human CD25 (Beckman Coulter, Krefeld, Germany) mab at a dilution of 1:1 and corresponding isotypes. The samples were analyzed by flow cytometry Beckman Coulter (Krefeld, Germany). Statistical analysis Results are expressed as mean±s.d. Characteristics of patients were compared by Student s t-test as appropriate. Variations in peripheral blood count, phenotype, and functional data were assessed by oneway analysis of variance with post hoc test by Scheffe for pair wise comparison. Each experimental condition was repeated three times. A Po.5 was considered significant. DISCLOSURE All the authors declare no conflict of interest. ACKNOWLEDGMENTS This work was funded by the Deutsche Bundesstiftung Umwelt grant (no. DBU 1611) for Scientific Research, Germany. REFERENCES 1. Fehervari Z, Sakaguchi S. Development and function of CD25+CD4+ regulatory T cells. Curr Opin Immunol 24; 16: Sakaguchi S, Sakaguchi N. Regulatory T cells in immunologic selftolerance and autoimmune disease. Int Rev Immunol 25; 24: Wood KJ, Sakaguchi S. Regulatory T cells in transplantation tolerance. Nat Rev Immunol 23; 3: Najafian N, Salama AD, Fedoseyeva EV et al. Enzyme-linked immunosorbent spot assay analysis of peripheral blood lymphocyte reactivity to donor HLA-DR peptides: potential novel assay for prediction of outcomes for renal transplant recipients. J Am Soc Nephrol 22; 13: Salama AD, Najafian N, Clarkson MR et al. Regulatory CD25+ T cells in human kidney transplant recipients. J Am Soc Nephrol 23; 14: Noris M, Casiraghi F, Todeschini M et al. Regulatory T cells and T cell depletion: role of immunosuppressive drugs. J Am Soc Nephrol 27; 18: Meloni F, Morosini M, Solari N et al. Peripheral CD4+ CD25+ Treg cell expansion in lung transplant recipients is not affected by calcineurin inhibitors. Int Immunopharmacol 26; 6: Calvo-Turrubiartes M, Romano-Moreno S, Garcia-Hernandez M et al. Quantitative analysis of regulatory T cells in kidney graft recipients: A relationship with calcineurin inhibitor level. Transpl Immunol 29; 21: Ekberg H, Tedesco-Silva H, Demirbas A et al. Reduced exposure to calcineurin inhibitors in renal transplantation. N Engl J Med 27; 357: Caproni M, Torchia D, Antiga E et al. The effects of tacrolimus ointment on regulatory T lymphocytes in atopic dermatitis. J Clin Immunol 26; 26: Khattri R, Cox T, Yasayko SA et al. An essential role for Scurfin in CD4+CD25+ T regulatory cells. Nat Immunol 23; 4: Wu Y, Borde M, Heissmeyer V et al. FOXP3 controls regulatory T cell function through cooperation with NFAT. Cell 26; 126: Kist-van Holthe JE, Gasser M, Womer K et al. Regulatory functions of alloreactive Th2 clones in human renal transplant recipients. Kidney Int 22; 62: Waaga AM, Gasser M, Kist-van Holthe JE et al. Regulatory functions of self-restricted MHC class II allopeptide-specific Th2 clones in vivo. J Clin Invest 21; 17: Webster AC, Woodroffe RC, Taylor RS et al. Tacrolimus versus ciclosporin as primary immunosuppression for kidney transplant recipients: meta-analysis and meta-regression of randomised trial data. BMJ 25; 331: Waaga AM, Gasser M, Laskowski I et al. Mechanisms of chronic rejection. Curr Opin Immunol 2; 12: Battaglia M, Stabilini A, Roncarolo MG. Rapamycin selectively expands CD4+CD25+FoxP3+ regulatory T cells. Blood 25; 15: Fontenot JD, Rasmussen JP, Williams LM et al. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity 25; 22: Ng WF, Duggan PJ, Ponchel F et al. Human CD4(+)CD25(+) cells: a naturally occurring population of regulatory T cells. Blood 21; 98: Sakaguchi S. Regulatory T cells: key controllers of immunologic self-tolerance. Cell 2; 11: Sakaguchi S, Ono M, Setoguchi R et al. Foxp3+ CD25+ CD4+ natural regulatory T cells in dominant self-tolerance and autoimmune disease. Immunol Rev 26; 212: Valencia X, Lipsky PE. CD4+CD25+FoxP3+ regulatory T cells in autoimmune diseases. Nat Clin Pract Rheumatol 27; 3: Meloni F, Vitulo P, Bianco AM et al. Regulatory CD4+CD25+ T cells in the peripheral blood of lung transplant recipients: correlation with transplant outcome. Transplantation 24; 77: Louis S, Braudeau C, Giral M et al. Contrasting CD25hiCD4+T cells/foxp3 patterns in chronic rejection and operational drug-free tolerance. Transplantation 26; 81: Bolton EM. Regulatory T cells in transplantation: natural or induced? Transplantation 25; 79: Waaga-Gasser AM, Grimm MR, Lutz J et al. Regulatory allospecific T cell clones abrogate chronic allograft rejection. J Am Soc Nephrol 29; 2: Baan CC, van der Mast BJ, Klepper M et al. Differential effect of calcineurin inhibitors, anti-cd25 antibodies and rapamycin on the induction of FOXP3 in human T cells. Transplantation 25; 8: Korczak-Kowalska G, Wierzbicki P, Bocian K et al. The influence of immuosuppressive therapy on the development of CD4+CD25+ T cells after renal transplantation. Transplant Proc 27; 39: Kidney International (211) 79,